On Friday, November 4, 2016, Juho Mäkinen <[email protected]> wrote:
> How do kubernetes users manage their manifest files? I'm asking because > we're transitioning into kubernetes and we are facinc several different > reasons why we would need some kind of templating engine to customise our > manifest files before they are imported into the cluster with kubectl. > > For example we are planning to run a separated clusters for production, > staging and testing. We would like that the application developers would > write a single manifest, which would then be applied first to testing, then > to staging and later to production. But to do this there are number of > reasons how we would need to customise the manifest files: > - Set externalIPs list depending on cluster > You talk about aws ebs later, and now about this. I'm aws you can just use a load balancer. We use a lb created by us with a service type node Port (instead of type load balancer). > - production will usually have more replicas than for example testing or > staging > We use a samson (github/zendesk/samson) for deployment and you can configure different replicas there. At deploy time, it will adjust the yaml on the fly > - Upgrade a service/prod to a new container image > We just update the yaml when we want this. we usually don't do it in all the projects at the same time, just in case. And each person responsible for each project does this. For example if some security update was done in our base images > - Services requiring persistent storage will most likely need for example > to have awsElasticBlockStore set dynamically. > > Why not use PVC? -- You received this message because you are subscribed to the Google Groups "Kubernetes user discussion and Q&A" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/kubernetes-users. For more options, visit https://groups.google.com/d/optout.
